RES EARCH

WOOD ENGINEERING Multiplex genome editing strategies for

concurrent wood trait improvement
Multiplex CRISPR editing of wood for sustainable Gene targets for multiplex genome editing
were identified using our established pre-
fiber production dictive model for monolignol biosynthesis
(13, 15, 22, 24). The model predicts the trans-
duction of quantitative relationships from
Daniel B. Sulis1,2,3, Xiao Jiang4, Chenmin Yang1,2,3, Barbara M. Marques2,3, Megan L. Matthews5,6,
gene transcript abundances to absolute en-
Zachary Miller4, Kai Lan4, Carlos Cofre-Vega2,3, Baoguang Liu1,2,3,7, Runkun Sun4,
zyme abundances, pathway metabolic fluxes,
Henry Sederoff2, Ryan G. Bing8, Xiaoyan Sun9, Cranos M. Williams3,5, Hasan Jameel4,
and 25 wood chemical and physical properties.
Richard Phillips4, Hou-min Chang4, Ilona Peszlen4, Yung-Yun Huang10, Wei Li11, Robert M. Kelly8,
Using the predictive model, we explored
Ronald R. Sederoff1,2,11, Vincent L. Chiang1,2,4,11, Rodolphe Barrangou1,3,12*, Jack P. Wang1,2,3,11*
69,123 sets of multigenic editing strategies
to reveal the extent to which individual
The domestication of forest trees for a more sustainable fiber bioeconomy has long been hindered wood properties can be modified through
by the complexity and plasticity of lignin, a biopolymer in wood that is recalcitrant to chemical multiplex editing of monolignol genes (Fig. 1,
and enzymatic degradation. Here, we show that multiplex CRISPR editing enables precise woody A and B). The 69,123 strategies encompass
feedstock design for combinatorial improvement of lignin composition and wood properties. By 48,831 gene combinations that represent all

Downloaded from https://www.science.org at University of Missouri Columbia on July 17, 2023

assessing every possible combination of 69,123 multigenic editing strategies for 21 lignin biosynthesis possible permutations of loss-of-function edit-
genes, we deduced seven different genome editing strategies targeting the concurrent alteration ing of up to six genes, and 20,292 gene com-
of up to six genes and produced 174 edited poplar variants. CRISPR editing increased the wood binations that target the combinatorial editing
carbohydrate-to-lignin ratio up to 228% that of wild type, leading to more-efficient fiber pulping. of up to three genes and/or overexpression of
The edited wood alleviates a major fiber-production bottleneck regardless of changes in tree growth one gene (data S1A). For each gene combina-
rate and could bring unprecedented operational efficiencies, bioeconomic opportunities, and tion, the model estimated the corresponding
environmental benefits. abundances of monolignol enzymes, pathway
metabolic fluxes, and 25 wood properties. Tar-

W
geting multiple genes expanded the range of
ood is the most abundant carbon bio- polymer that cross-links with cellulose and phenotypic variation attainable by genome
mass on Earth and the major source hemicelluloses in the secondary cell walls editing compared with single-gene edits (Fig.
of sustainable green fibers (1). Glob- of vascular plants (8, 9). The polymer in an- 1C and data S1B). For example, multigenic
ally, ≈315 billion metric tons of car- giosperm wood is formed by free radical editing could reduce lignin content to 50.7%
bon are stored as wood, representing polymerization of three major monolignol of the wild-type level, whereas single-gene
≈57% of the biogenic carbon sink (1). The bio- precursors, 4-coumaryl, coniferyl, and sinapyl edits only reduced lignin content to 61.3% of
mass supplies >170 million metric tons of alcohols, which form the p-hydroxyphenyl the wild-type level (Fig. 1C and data S1B).
virgin fibers annually (2) to meet the growing (H), guaiacyl (G), and syringyl (S) units in Wood with low lignin content and a high
demand for renewable tissue, paper, pack- lignin (10–12). The monolignols are biosyn- syringyl-to-guaiacyl (S/G) ratio is ideal for
aging, textile, and other fiber products, in- thesized from phenylalanine through a se- fiber production (25). We mined all 69,123
cluding structural materials (3). Despite the ries of enzymatic reactions in a metabolic strategies to identify gene combinations that
importance of wood fibers, their production grid consisting of at least 11 enzyme families are predicted to reduce lignin content by at
has remained largely limited to undomes- and 24 metabolites (13). The pathway is me- least 15% and increase the carbohydrate-to-
ticated forest trees with often suboptimal diated by hundreds of regulatory influences, lignin (C/L) ratio by at least 200% and that
wood properties that hamper production encompassing transcriptional (14), transla- have a higher S/G ratio than wild type. The
efficiency. The propensity for efficient isola- tional (15), and posttranslational regulations C/L ratio is an indicator of the potential max-
tion of cellulosic fibers from wood is largely (16, 17); enzyme–enzyme interactions (18–20); imum cellulosic yield for wood fiber (13). The
determined by the content and composi- and metabolic regulations (21, 22). More strategies must also have predicted growth
tion of lignin (4–6), one of three major com- than five decades of research have exten- characteristics (e.g., tree height) that are com-
ponents of wood (7). Lignin is a phenolic sively investigated the individual components parable to (>75%) or exceed those of wild-type
of lignin biosynthesis and determined the controls. Of the 69,123 strategies, only 347
effects of their perturbation on lignin con- (0.5%) matched the aforementioned criteria in
1
TreeCo, Raleigh, NC, USA. 2Department of Forestry and tent and composition in diverse plant spe- lignin content, C/L and S/G ratios, and tree
Environmental Resources, North Carolina State University,
Raleigh, NC, USA. 3NC Plant Sciences Initiative, North cies (6, 13, 23). However, these efforts have growth (Fig. 1B and data S1C), highlighting
Carolina State University, Raleigh, NC, USA. 4Department predominantly focused on the modification the need to mine such strategies to practically
of Forest Biomaterials, North Carolina State University, of single genes or gene families, whereas the test the most-promising combinations. All 347
Raleigh, NC, USA. 5Department of Electrical and Computer
Engineering, North Carolina State University, Raleigh, NC,
combinatorial effects of multigenic pertur- strategies target at least two monolignol genes,
USA. 6Department of Civil and Environmental Engineering, bations have remained elusive. Here, we show and 99.7% of the strategies target at least three
University of Illinois at Urbana-Champaign, Champaign, IL, that strategic multiplex CRISPR editing of genes (Fig. 1D and data S1C). The number of
USA. 7Department of Forestry, Beihua University, Jilin, China.
8
Department of Chemical and Biomolecular Engineering, North
monolignol biosynthesis genes improves wood target genes affirms the need for a multigenic
Carolina State University, Raleigh, NC, USA. 9Molecular properties beyond what can be achieved by approach to improve fiber traits in Populus
Education, Technology and Research Innovation Center editing single genes or gene families and de- trichocarpa. Lignin gene families that most
(METRIC), North Carolina State University, Raleigh, NC, USA.
10 bottlenecks a key operational constraint in frequently appeared in these 347 strategies
Department of Operations Research, North Carolina State
University, Raleigh, NC, USA. 11State Key Laboratory of Tree industrial pulp mills. These improvements sub- were C3H (coumarate 3-hydroxylase), CCoAOMT
Genetics and Breeding, Northeast Forestry University, Harbin, stantially increase fiber production capacity (caffeoyl-CoA O-methyltransferase), AldOMT
China. 12Department of Food, Bioprocessing and Nutrition while reducing the global warming potential (aldehyde O-methyltransferase), PAL (phenyl-
Sciences, North Carolina State University, Raleigh, NC, USA.
*Corresponding author. Email: rbarran@ncsu.edu (R.B.); of pulp mills, leading to a more sustainable alanine ammonia-lyase), C4H (cinnamate
jpwang@ncsu.edu (J.P.W.) and efficient fiber bioeconomy. 4-hydroxylase), and CAD (cinnamyl alcohol

Sulis et al., Science 381, 216–221 (2023) 14 July 2023 1 of 6

RES EARCH | R E S E A R C H A R T I C L E

dehydrogenase) (Fig. 1E). We then performed

sensitivity analysis for these gene families to
examine the robustness of fiber trait improve-
ments in response to varying transcript abun-
dances of nontargeted monolignol genes
(Fig. 1, F to I; fig. S1; and data S2 and S3)
and selected seven strategies for editing in
P. trichocarpa. The seven strategies (table S1)
encompassed various numbers of target genes
(from three to six) and were selected on the
basis of the extent and robustness of predicted
improvement in fiber traits (reduced lignin,
increased S/G and C/L ratios, and good growth)
(data S2 and S3). The selected strategies
showed a predicted reduction in lignin con-
tent of up to 35% compared with wild type
(Fig. 1F and data S3). S/G and C/L ratios in-

Downloaded from https://www.science.org at University of Missouri Columbia on July 17, 2023

creased up to 248 and 215%, respectively (Fig.
1, G and H, and data S3), with no change in
tree height (Fig. 1I and data S3).

Generating genetic diversity in poplar

using CRISPR
To test the seven strategies in planta for their
modulation of wood properties for fiber pro-
duction, a multiplex CRISPR construct (26)
was assembled for each strategy and delivered
into P. trichocarpa using Agrobacterium
tumefaciens (27) (figs. S2 to S4, data S4,
and tables S2 and S3). We generated 174 inde-
pendent lines of multiplex CRISPR–edited
P. trichocarpa, including 78 lines that target
the concurrent editing of three monolignol
genes, 20 lines that target four genes, 41 lines
that target five genes, and 35 lines that target
six genes. The edited lines exhibited varying
extents of loss-of-function mutations of the
target genes (Fig. 2A, fig. S5, and data S5).
Biallelic loss-of-function editing of all target
genes was obtained in P. trichocarpa for strat-
egies that target the concurrent editing of
three or four genes (Fig. 2A, fig. S5, and data
S5). For strategies that target five or six genes,
none of the 76 edited lines showed complete
loss-of-function editing of both alleles in all
target genes. Nonetheless, we identified sev-
eral lines (e.g., J-7 and K-28) that harbor sub-
stantial editing of most target genes (fig. S5
and data S5). The frequency of biallelic edits
varied between target genes, ranging from
86.1% for PtrPAL4 and PtrPAL5 to 9.3% for
PtrCCoAOMT2 (Fig. 2B). Most edits consisted
of small insertions and deletions (INDELs)
of 1 to 3 nucleotides (nt) in the immediate
vicinity of the targeted cleavage site (Fig. 2, C
Fig. 1. Identification of multiplex CRISPR editing strategies to improve wood fiber traits. and D, and fig. S6). However, deletions of up
(A) Three-dimensional scatterplot of predicted lignin content, S/G ratio, and tree growth (stem to 19 nt were also occasionally observed (Fig.
diameter) for 69,123 genome editing strategies. (B) Venn diagram of the 69,123 strategies showing 2D). No off-target edits were detected in our
the relationships between predicted lignin content, S/G ratio, C/L ratio, and tree growth (height). edited lines (table S4). We used RNA sequencing
(C) Distribution of lignin content in single-gene and multigene editing strategies. (D) The total number to evaluate two potential xylem-expressing off-
of target genes in the selected 347 strategies. (E) The frequency of each monolignol gene in the target genomic loci in 12 edited P. trichocarpa
347 strategies. (F to I) Global sensitivity analysis of wild-type (a), single-gene editing (b to d), and lines (data S6). The absence of detectable off-
multigene editing (e to k) strategies for lignin content (F), S/G ratio (G), C/L ratio (H), and tree growth target editing suggests that genome editing
(height) (I). For additional information, see fig. S1. is highly specific in P. trichocarpa. The broad

Sulis et al., Science 381, 216–221 (2023) 14 July 2023 2 of 6

RES EARCH | R E S E A R C H A R T I C L E

Fig. 2. Genetic diversity in multiplex CRISPR–edited P. trichocarpa. mutations in PtrPAL2 (C) and PtrPAL4 and PtrPAL5 (D). Stem cross sections of Downloaded from https://www.science.org at University of Missouri Columbia on July 17, 2023
(A) Percentage INDELs of each target gene in the 58 CRISPR-edited P. trichocarpa wild-type (E) and CRISPR-edited H-19-3 trees (F), with the H-19-3 tree showing
lines targeting the concurrent editing of three monolignol genes (PtrPAL2, PtrPAL4, distinct red coloration of the xylem. (G) Six-month-old greenhouse-grown
and PtrPAL5). (B) Average percentage INDELs of the target monolignol genes in CRISPR-edited and wild-type P. trichocarpa. (H) Harvested stem segments from
the 174 edited P. trichocarpa lines. (C and D) Most-frequent CRISPR-Cas–mediated CRISPR-edited and wild-type P. trichocarpa.

variation in the editing profiles of target genes Phenotypic characterization of multiplex tains 20.9% lignin and has a C/L ratio of 3.0
in our 174 edited lines created genetic diversity CRISPR–edited poplar (data S7). In the multiplex-edited wood, lignin
in monolignol biosynthesis not present in Multiplex editing of monolignol genes in pop- content is reduced by up to 49.1%, and the C/L
nature (Fig. 2, E to H). Such genetic diversity lar substantially altered the chemical and ratio increased up to 228%, compared with wild
enables an exploration of how multigenic path- physical properties of the wood (Fig. 3, A to D; type (Fig. 3, E and F, and data S7, determined
way perturbations can combinatorially regulate fig. S7; and data S7 to S10), as predicted by the as averages across 65 lines). Two-dimensional
wood formation and wood utilization, as well lignin model (fig. S8 and table S5). Stem wood nuclear magnetic resonance spectroscopy
as enhance genetic diversity. of 6-month-old wild-type P. trichocarpa con- (2D NMR) (Fig. 3, G to J, and figs. S9 to S11)

Sulis et al., Science 381, 216–221 (2023) 14 July 2023 3 of 6

RES EARCH | R E S E A R C H A R T I C L E

Downloaded from https://www.science.org at University of Missouri Columbia on July 17, 2023

Fig. 3. Phenotypic variation in CRISPR-edited and wild-type P. trichocarpa. traits. Error bars represent one standard error of up to three biological replicates.
(A to D) Distribution of lignin content (A), C/L ratio (B), S/G ratio (C), and tree growth (G to J) 2D NMR 13C-1H (heteronuclear single quantum coherence) correlation
(height) (D) in CRISPR-edited and wild-type P. trichocarpa. Blue and red colors spectra (side-chain and aromatic regions) of wild type [(G) and (H)] and the
stand for CRISPR-edited lines and wild-type control measurements, respectively. G-27 edited line (see fig. S8 for other edited lines). Contours in these regions
Pink represents the overlapped regions of the blue (CRISPR-edited lines) and red were used to estimate the distribution of lignin interunit linkages and lignin
(wild-type controls) bars. Lignin, C/L ratio, and growth data are from E, F, G, H, I, J, composition, namely S/G/H ratios, as well as p-hydroxybenzoate (PB). ppm,
and K constructs. S/G ratio data are from E, F, and G constructs. Average lignin parts per million. (K) Quantitative relationship between lignin content and
content (E) and C/L ratio (F) of selected CRISPR-edited lines with improved fiber wood volume in CRISPR-edited and wild-type P. trichocarpa.

revealed that multigenic strategies also modu- lines (E-9 and G-25), which had reduced wood Projected industrial benefits of
lated lignin composition, increasing the S/G elasticity (fig. S7 and data S9). No change in debottlenecking fiber production
ratio from 2.7 in wild type to as high as 4.0 in wood density greater than 15.1% (detection Kraft pulping is the dominant process for in-
the edited lines (Fig. 3C and data S8). As limit due to the large technical and biological dustrial wood fiber production (28–30). To
expected, lines harboring single-gene edits variation; see materials and methods in the understand the techno-economic impacts of
showed either no reduction or a mild reduc- supplementary materials) of the wild-type CRISPR-edited wood on kraft pulping, we
tion in lignin content (0.1 to 12.6%) (e.g., lines level was observed in the edited P. trichocarpa used a Carolina Pulp and Paper (CPP) mill
G-7, I-17, and K-9; fig. S12 and data S5 and S7). (data S9). Several CRISPR-edited lines with model (fig. S13) based on an existing industrial
The most significant lignin reductions were significant reductions in lignin content showed Brazilian pulp mill with an annual production
observed in edited trees that harbor four to six a reduction in tree growth (e.g., I-1, J-25, and of 1.24 million tons of pulp. Pulp yield and
gene edits, but strategies that targeted three K-6) (Fig. 3, D and K; figs. S7 and S12; and data production efficiency may be significantly in-
genes also showed significant lignin reduc- S7 and S10). Some multiplex genome editing creased by multiplex editing of monolignol
tions of up to 32% (Fig. 3E, fig. S12, and data strategies can circumvent certain shortcom- genes (Fig. 4, A and B), as evidenced by mi-
S7). Edited lines with significant changes in ings inherent to typical single-knockout strat- cropulping wood with varying lignin content
wood chemical properties predominantly har- egies that often have undesirable phenotypic and composition. Reducing lignin content
bored a single-allele loss-of-function mutation effects on combinations of tree characteristics proportionally increases pulp yield and could
in PtrC3H3 combined with the mutation of and wood properties. Moving forward, it will reduce the usage of pulping chemicals (Fig. 4,
either one or both alleles of PtrAldOMT2 and be necessary to monitor the impact of the A and B; figs. S14 to S17; and tables S6 to S8)
PtrCAD1 (fig. S12). Wood elasticity was not genome editing outcomes on the tree pheno- that are undesirable from an environmental
significantly different between wild-type and typic properties and traits relevant to the in- standpoint. Furthermore, low-lignin wood
genome-edited P. trichocarpa, except for two dustry in long-term field trials. could reduce the solid content of black liquor,

Sulis et al., Science 381, 216–221 (2023) 14 July 2023 4 of 6

RES EARCH | R E S E A R C H A R T I C L E

Downloaded from https://www.science.org at University of Missouri Columbia on July 17, 2023

Fig. 4. Techno-economic analysis and global warming potential of CRISPR-edited wood in industrial kraft pulping. (A) Correlation between lignin content
and pulp yield in CRISPR-edited and wild-type P. trichocarpa. (B) Paper disks produced from CRISPR-edited P. trichocarpa. (C) The impact of varying lignin
content and S/G ratio on the incremental production potential of an industrial kraft pulp mill. (D and E) Operational capacity of different components of an
industrial kraft pulp mill using wood with 28% lignin (D) or 16% lignin (E) at an S/G ratio of 2.8. (F) Total life-cycle GWP of producing pulp from 1 BDMT log.
The scenarios include varied S/G ratios (2.8, 4.0, or 6.0), lignin content (28 or 16%), and production modes (constant or incremental).

thereby debottlenecking the recovery boiler, Sustainability impact of optimized technologies for tree improvement, which may
arguably the most crucial and rate-limiting fiber production be further corroborated in field trials in rele-
energetic component of pulp mills (Fig. 4, C The pulp and paper industry is a major con- vant environmental conditions. This study
to E, and fig. S18). The debottlenecking then tributor to greenhouse gas emissions. The US illustrates how strategic multiplex editing to
enables an incremental production potential Environmental Protection Agency reported alter wood composition could enable more-
of the pulp mill by up to 40% (Fig. 4, C to E, 35 million metric tons of CO2 equivalent emis- sustainable fiber production with remarkable
and fig. S18). Additionally, increasing the C/L sions from the pulp and paper industry in the operational efficiencies, bioeconomic value cre-
ratio in wood means that less biomass is US in 2020 (31). Globally, the annual direct CO2 ation, and tangible environmental benefits.
required to provide the same amount of cel- emissions from pulp and paper production
lulose (fig. S17). The benefits of increasing the reached 168 million metric tons (32). We as- REFERENCES AND NOTES
S/G ratio encompassed the reduction of pulp- sessed the impacts of CRISPR-edited wood on 1. Y. M. Bar-On, R. Phillips, R. Milo, Proc. Natl. Acad. Sci. U.S.A.
ing chemicals and energy savings on chemical the carbon footprint of pulp production. Life- 115, 6506–6511 (2018).
recovery (figs. S16 and S18). Tremendous fi- cycle global warming potential (GWP) was 2. RISI, Fastmarkets, “RISI annual review of global pulp and paper
statistics” (2022); https://www.risiinfo.com/product/annual-
nancial benefits may be obtained by reducing estimated for pulp production from 1 bone dry review-of-global-pulp-paper-statistics.
lignin from 28% to 16% and by increasing the metric ton (BDMT) of CRISPR-edited wood 3. S. Xiao et al., Science 374, 465–471 (2021).
S/G ratio from 2.8 to 4.0 (figs. S19 to S21). Such and compared with that of wild-type wood. 4. J. Ralph et al., Science 277, 235–239 (1997).
5. R. Vanholme et al., Science 341, 1103–1106 (2013).
changes are predicted to increase the net pres- For a pulp mill powered by natural gas, reduc- 6. F. Chen, R. A. Dixon, Nat. Biotechnol. 25, 759–761
ent value (at a 14% hurdle rate) from −$359.3 ing lignin content from 28% to 16% decreases (2007).
million to $1047.1 million USD for pulp mills the GWP by up to 8% (Fig. 4F, figs. S22 to S26, 7. R. C. Pettersen, Adv. Chem. Ser. 207, 57–126
(1984).
using natural gas as supplementary energy, and tables S9 and S10), and increasing the S/G 8. Q. Li et al., Plant Biotechnol. J. 12, 1174–1192 (2014).
and from $269.2 million to $1856.9 million ratio from 2.8 to 4.0 leads to a further 10% 9. K. V. Sarkanen, C. H. Ludwig, Lignins: Occurrence, Formation,
USD for mills using biomass as supplemen- reduction in the GWP per BDMT of logs pro- Structure and Reactions (Wiley-Interscience, 1971).
10. W. Boerjan, J. Ralph, M. Baucher, Annu. Rev. Plant Biol. 54,
tary energy (figs. S19, C and D, S20, and S21). cessed (Fig. 4F, figs. S22 to S26, and tables S9 519–546 (2003).
Notably, some of these benefits are achieved and S10). Altogether, incremental pulp produc- 11. R. Vanholme, B. Demedts, K. Morreel, J. Ralph, W. Boerjan,
even in cases of biomass reduction because tion using edited wood with reduced lignin Plant Physiol. 153, 895–905 (2010).
12. K. Freudenberg, Science 148, 595–600 (1965).
of industrial operational efficiency gains (figs. and an increased S/G ratio could reduce GWP
13. J. P. Wang et al., Nat. Commun. 9, 1579 (2018).
S20 and S21). Although the magnitude of the by up to 20% (Fig. 4F, figs. S22 to S26, and 14. H. Chen et al., Plant Cell 31, 602–626 (2019).
financial gains is forecasted, accuracy will tables S9 and S10), providing substantial bene- 15. M. L. Matthews, J. P. Wang, R. Sederoff, V. L. Chiang,
eventually be determined by the physiolog- fits to environmental conservation and climate C. M. Williams, PLOS Comput. Biol. 16, e1007197
(2020).
ical attributes of the harvested trees at in- change mitigation. These results provide in- 16. J. P. Wang et al., Proc. Natl. Acad. Sci. U.S.A. 112, 8481–8486
dustrial scale. sights into the potential of genome editing (2015).

Sulis et al., Science 381, 216–221 (2023) 14 July 2023 5 of 6

RES EARCH | R E S E A R C H A R T I C L E

17. D. B. Sulis, J. P. Wang, Front. Plant Sci. 11, 914 31. US Environmental Protection Agency, Greenhouse Gas D.B.S., J.P.W., X.J., K.L., B.M.M., C.-C.V., and R.B. Competing
(2020). Reporting Program (GHGRP) (2021); https://www.epa.gov/ interests: R.B. and J.P.W. are shareholders of TreeCo. All other
18. J. P. Wang, B. Liu, Y. Sun, V. L. Chiang, R. R. Sederoff, Front. ghgreporting/ghgrp-pulp-and-paper#map-facilities. authors declare that they have no competing interests. Data and
Plant Sci. 9, 1942 (2019). 32. International Energy Agency, Tracking Industry (2020); https://www. materials availability: All data are available in the main text or the
19. H. C. Chen et al., Plant Cell 26, 876–893 (2014). iea.org/reports/tracking-industry-2020 [accessed 18 July 2021]. supplementary materials. License information: Copyright © 2023
20. H.-C. Chen et al., Proc. Natl. Acad. Sci. U.S.A. 108, the authors, some rights reserved; exclusive licensee American
21253–21258 (2011). Association for the Advancement of Science. No claim to original
ACKN OWLED GMEN TS
21. J. P. Wang et al., Plant Cell 26, 894–914 (2014). US government works. https://www.science.org/about/science-
We thank K. Osakabe for providing the multiplex genome editing licenses-journal-article-reuse
22. J. P. Wang et al., Curr. Opin. Biotechnol. 56, 187–192
vectors and associated protocols for constructs assembly used
(2019).
in this paper. Funding: This work was supported by National Institute SUPPLEMENTARY MATERIALS
23. A. J. Ragauskas et al., Science 344, 1246843 of Food and Agriculture of the US Department of Agriculture–
(2014). science.org/doi/10.1126/science.add4514
Agriculture and Food Research Initiative grant 2018-67021-27716;
24. M. L. Matthews, J. P. Wang, R. Sederoff, V. L. Chiang, Materials and Methods
North Carolina State University Chancellor’s Innovation Fund grant
C. M. Williams, Comput. Struct. Biotechnol. J. 19, 168–182 Supplementary Text
190549MA; North Carolina State University Goodnight Early Career
(2020). Figs. S1 to S26
Innovator Award (to J.P.W.); US National Science Foundation Small
25. R. B. Santos, E. A. Capanema, M. Y. Balakshin, H.-M. Chang, Tables S1 to S10
Business Technology Transfer Program grant 2044721; Cooperative
H. Jameel, BioResources 6, 3623–3637 (2011). References (33–40)
State Research Service of the US Department of Agriculture grant
26. R. Ueta et al., Sci. Rep. 7, 507 (2017). Data S1 to S10
NCZ04214; and North Carolina Specialty Crop Block Grants 19-019-
27. J. Song, S. Lu, Z. Z. Chen, R. Lourenco, V. L. Chiang, Plant Cell MDAR Reproducibility Checklist
4018, 19-092-4012, and 20-070-4013. Author contributions:
Physiol. 47, 1582–1589 (2006). Conceptualization: J.P.W., V.L.C., R.R.S., and R.B. Methodology: D.B.S., View/request a protocol for this paper from Bio-protocol.
28. P. J. Kleppe, Tappi 53, 35–47 (1970). X.J., C.Y., M.L.M., C.M.W., H.J., R.P., H.-m.C., I.P., Y.-Y.H., and W.L.

Downloaded from https://www.science.org at University of Missouri Columbia on July 17, 2023

29. J. Gierer, Wood Sci. Technol. 14, 241–266 (1980). Investigation: D.B.S., X.J., C.Y., B.M.M., Z.M., K.L., C.-C.V., B.L., R.S., H.S., Submitted 12 June 2022; resubmitted 20 November 2022
30. N. Bijok, J. Fiskari, R. R. Gustafson, V. Alopaeus, Chem. Eng. J. R.G.B., and X.S. Funding acquisition: J.P.W., R.M.K., and R.B. Supervision: Accepted 16 May 2023
438, 135548 (2022). J.P.W., R.B., C.M.W., V.L.C., H.J., R.P., H.-m.C., I.P., and R.R.S. Writing: 10.1126/science.add4514

Sulis et al., Science 381, 216–221 (2023) 14 July 2023 6 of 6

 Multiplex CRISPR editing of wood for sustainable fiber production
Daniel B. Sulis, Xiao Jiang, Chenmin Yang, Barbara M. Marques, Megan L. Matthews, Zachary Miller, Kai Lan, Carlos
Cofre-Vega, Baoguang Liu, Runkun Sun, Henry Sederoff, Ryan G. Bing, Xiaoyan Sun, Cranos M. Williams, Hasan Jameel,
Richard Phillips, Hou-min Chang, Ilona Peszlen, Yung-Yun Huang, Wei Li, Robert M. Kelly, Ronald R. Sederoff, Vincent L.
Chiang, Rodolphe Barrangou, and Jack P. Wang

Science, 381 (6654), .

DOI: 10.1126/science.add4514

Downloaded from https://www.science.org at University of Missouri Columbia on July 17, 2023

Editor’s summary
Trees provide an important natural resource, but breeding for optimal wood properties is time consuming and hindered
by the complexity of tree genetics and diversity. Sulis et al. show that CRISPR technologies can be readily deployed
to enhance wood properties and augment the sustainability of forest trees (see the Perspective by Zuin Zeidler). The
authors generated multiplexed genetic alterations modifying wood composition in poplar with more desirable traits for
fiber pulping and lower carbon emissions. This work demonstrates that genome editing can be harnessed for breeding
more efficient trees, which will provide timely opportunities for sustainable forestry and a more efficient bioeconomy. —
DJ

View the article online

https://www.science.org/doi/10.1126/science.add4514
Permissions
https://www.science.org/help/reprints-and-permissions

Use of this article is subject to the Terms of service

Science (ISSN ) is published by the American Association for the Advancement of Science. 1200 New York Avenue NW, Washington, DC
20005. The title Science is a registered trademark of AAAS.
Copyright © 2023 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. No claim
to original U.S. Government Works